Posted: Sep 05, 2010 2:25 pm
by jerome
Defender's #8 post: 2959 words - Mr.Samsa

Knocked off?

Even by the standards of the debate it has been a long time, while I took s much needed holiday away from the net to run a games convention. After the frenetic and extremely valuable burst of activity on the commentary thread while we analysed Dr Colvin’s paper I hope I can be forgiven!
In his last post Campemon gave a superb and concise description of our attempts with GrahamH and Twistor59 to test the claims made by Colvin in the JSPR. I created the control files myself, and then gave them blind to he other three – and my method for creating the control files was extremely simple – I banged on furniture! One file was created by my ex-wife bouncing upstairs on the floor and me recording the noise in the basement, another dropped from the analysis was an abrasion; we dropped this simply because it did not sound like a rap in any sense, though it certainly showed the purportedly characteristic waveform.

Dr. Barrie Colvin wrote:
“There appear to be reasonable grounds for concluding that the unexplained rapping effect produced at various apparent poltergeist cases in a variety of countries exhibit an unusual acoustic waveform pattern, characterized by a relatively slow rise to maximum amplitude, followed by an equally slow decline in amplitude.”

We only tested attack time. I am not convinced that it would have made much difference if we had tested the decline as well: it was Twistor59 who meticulously developed a set of parameters for what constitutes “attack time” and applied them to the data; this numerical analysis proved invaluable. I fully agree with Campermon when he writes

Campermon wrote:
I would suggest that if Colvin were to show that poltergeist raps were ‘anomalous’ then he should, as we have done in a half arsed manner (‘a bit Heath-Robinson’ as Twistor put it!), make quantitative measurements on these raps and compare them to many (100+) self produced raps. If the poltergeist raps are indeed different, then they should stick out like a ‘dog log in a swimming pool’.

I am intrigued by Campermon’s suggestion of aliasing: it should have been addressed in the paper; though to be fair to Dr Colvin, he s carefully conducting further research. I have invited the main contributors to our “beer mat physics” experiments and one other authority on acoustics physics to collaborate on a paper critiquing the claims, which hopefully we can put together quite quickly. It was certainly extremely useful to bring our full critical attention to bear on the paper, and I think I can safely accept that the paper I brought to the debate has been demolished by the experiments I instigated and the critical thinking of Campermon and the commentary participants. I am confident that our critique will be published; since Honorton 1981 the parapsychological journals have published negative findings almost irrespective of the quality of the paper, to prevent file drawer effects (most important in something like the Ganzfeld meta-analyses). So we have made what I feel is a very important contribution to the field of parapsychology.

How does it work?

Campermon has not really addressed why some waveforms (NOT just ‘poltergeist’ ones!) acquire this odd slow attack/decline pattern: which is odd as he is the sceptic and also his suggestion of abrasion (which was a brilliant one, and certainly correct in that rubbing two rough surfaces together generate the waveform in question, just sounds wrong) led me to decide to run the simple tests we used. I will therefore address it: please note that what I know about acoustic physics could be written on the back of my fingernails, so please feel free to correct me in the commentary thread!

I think two factors are involved, a) the presence of blank wall space that ‘reflects’ the sounds, creating a sustained rise – but I cant be sure that is a factor, I just think it possible b) the transmission of the sounds through two mediums (as demonstrated by the wav file of my wife bouncing above me through the floor). In most cases my final success in replicating the noises came from hitting an object some distance from the mc stand which rested on my desk – I suspect but do not know that the varying rates of propagation of the sound waves through two different mediums (the air and the desk) led to attenuation of the sound and the waveform characteristic displayed.

In fact the first reason I suspected this was the comments on my blog Polterwotsit when the commenter JG wrote
JG wrote: There is one thing I don’t understand though. The ‘normal’ sounds are indeed what you would expect if you hit something with nothing but a short stretch of air between you and the tape recorder. However, in poltergeist cases, aren’t the sounds usually said to come from inside or behind walls or furniture? If so, there is at least one layer of wood or plaster, etc between the origin of the sound and the air. This would have the effect of damping down the higher frequencies in the sound. Since the lower frequencies of a percussive sound take longer to reach full intensity, isn’t the slower build up overall what we might expect? Thus, the sound intensity build up would be normal at the point of origin of the sound but altered, with a lower build up and higher frequencies missing, once it had gone through a layer of wood. The same affect would occur if someone actually hit the other side of a wall from the room where the tape recorder was positioned.

Just to clarify my previous comments: when sound waves move through a solid, they split into two modes – transverse and longitudinal. These travel at different speeds (one approximately twice the speed of the other). Thus not all the sound will arrive together, so making the build up to peak intensity slower than if it was travelling purely through air. So, if the rap goes through a wall or item of furniture, you would expect a slower build up in sound intensity compared to simply going through air.

You can try this as an experiment by comparing someone hitting a wall inside a room and then again outside.

Absolutely JG, you hit the nail on the head, and my slow response to your comments has been caused by my testing this with the folks here!

And I would have gotten away with it…

Dr Colvin’s paper was very interesting, and I’m glad we dedicated so much time to it. Of course it proved worthless to my case: I am even now dedicating words where I should be making a positive case to analysing a failed line of argument: but let’s face it, if I had never brought it to the table we would never have critiqued it, and the critique is to my mind as valuable as the original claims.

We simply don’t know exactly where the microphone was positioned in most of these cases; however further acoustic analysis should perhaps use the cases where the filming of the person holding the microphone allows for a better understanding of the condition under which they were recorded. Was the mic on a table? Held in the air? On October 31st a Radio Four broadcast I was present during the recording of will feature some purportedly anomalous knocks (and hopefully make you laugh, you will understand when it is aired I promise!) where I happen to know the conditions under which it was recorded. I look forward to analysing those sounds, but don’t expect to find anything unusual.

Now of course some people will inevitably cite all this as proof that poltergeist cases are nonsense: that is clearly untrue. What we have demonstrated is that the sounds that were claimed as anomalous are not anomalous after all, or at least were replicable without any real effort by very simple techniques perfectly consistent with the conditions in which they were recorded – banging on walls or furniture!

This makes me wonder about the assumptions that underlie the original claim, that poltergeists make sounds somehow different to ‘normal’ sounds. So I looked carefully at what Dr Colvin proposed; he suggests the sounds may originate from within the substance of a material, say a table for example, from which the raps appear to emanate. This is indeed consistent with the waveform characteristics – unfortunately so is someone knocking on an adjoining wall, or even within the same room.

Dr Colvin has suggested experiments with “sitters” at séances designed to cause physical phenomena trying to create anomalous sounds. Oddly I don’t think Colvin holds to the spirit hypothesis; he is working on the assumption that poltergeists are a result of an unknown living agent performing some form of psychokinesis as far as I understand. I may be wrong! However if the waveforms were anomalous a whole new technique for investigating PK would open up: Colvin is working within a naturalistic paradigm, where he believes that poltergeists work within the laws of physics – as we would expect from a parapsychologist. He does not claim in any way that these sounds are “supernatural” – merely that they are consistent with having originated in a physical change within the structure of an object. Sure, that is true: however as I hope we have demonstrated they are also consistent with plenty of other options. One, banging on the walls, makes me think…

Back to Andover!

It is an important precept of rational discourse that one attacks the ideas, not the person holding them. However in this case I was curious about how Dr Colvin came to look for unusual characteristics in the rapping phenomena (and secondly, why the paper was not subjected to some simple tests and greater critique in peer review, but I am afraid that lacking is normal in many fields.). It was a great idea: but why did Dr Colvin develop it?
The answer appears to me simple: back in ’74 he investigated a poltergeist case in Andover, England. We have referenced it and discussed it in passing earlier in the debate (for a brief discussion read the blog post here -- ... r-pol.html )

What struck me as interesting was that the phenomena there chiefly occurred in a room that was adjacent to another house in a semi-detached property; because the family wanted no publicity, delaying the publication of Colvin’s paper on that case for over twenty years. However, it remains possible the neighbours were involved in faking the phenomenon – people being people, and the neighbours being unaware one assumes of the investigation, it my take a television account of the case before the neighbours come forward if they were involved, if they do even then..

Against that possibility is the original newspaper story featuring the ghost – Mystery ghost making their lives a misery. Andover Advertiser (26 April, 1974) which means that if the neighbours were hoaxing they had good reason to know what was occurring next door.

(The case had been covered in Dr Terry White’s The Sceptical Occultist, in the mid-1990’s but pseudonyms were used.) The family still want no publicity – and so while I have seen plenty of evidence that the case occurred and developed as Colvin reports (photographs, an unpublished manuscript of Colvin’s write up on the case that accidentally came in to my possession fifteen years ago, and footage and recordings that he shared at the SPR Study Day on Poltergeists) what led Colvin to eschew the neighbours as a likely source, and look for something odd in the sounds?
It certainly was an explanation considered at the time – in fact the very first explanation that occurred to the family. I have edited the girl’s names down to initials to protect their identities – the internet is rather wider circulation than the JSPR, though these may be pseudonyms anyway, I have not checked.

Colvin, JSPR, vol. 72; No.890; January 2008 wrote:
The sounds were first heard by T and M during the Easter weekend. The two girls explained that light tapping sounds were first heard when they lay quietly in bed on Good Friday, 12 April. At first they thought that the sounds were being made by someone next door, especially since the tapping sounds always emanated from the wall adjacent to T's bed. However, the girls soon became convinced that this was not the real explanation because they obtained knocking replies to questions which Theresa whispered so softly that her sister M, lying in the adjacent bed, could hardly hear. They felt that in these circumstances no one in the adjacent house could have heard them talking. M and T explained that at first they were not at all troubled by these events, which occurred nightly. In fact, they positively enjoyed the experience and looked forward to going to bed each evening.

So yes, that was the very first possibility considered. However what made the case startling was the fact that the ‘poltergeist’ communicated with a simple alphabetical code of knocks – A was 1, B was 2, and so on. I have heard recordings (at the study Day) of these coded messages, and they were delivered swiftly and without error; something I would find impossible I fear, but which is not beyond the sound of possibility for someone willing to train themselves in the skill I expect. Dr Colvin now tried an interesting experiment

Colvin, JSPR, vol. 72; No.890; January 2008 wrote:
Having prepared a total of 40 shuffled cards, each one depicting a number from 1 to 10 (i.e. 4 sets), I carried out several runs in the following manner: Theresa was lying on her bed, face down, with her head turned slightly towards the wall. Mrs Andrews and Kevin sat on Maria's bed and I stood close to the window, facing Theresa. I picked a card at random and showed it to everyone in the room, including Theresa, without actually stating what the number was. I held it up so that it faced the wall and asked Eric to tap out the number on the card. Four slow but definite raps were heard to come from the wall. The last rap was louder than the others. After the fourth rap, I continued to hold the card in place in case any further raps were forthcoming. When it was clear that the rapping had finished, I confirmed to Eric that the answer was correct. This procedure was repeated five times and on each occasion Eric produced the correct number of raps.

The tests were then continued in a slightly different manner. I picked a card out of the pack and, ensuring that Theresa was facing the wall, looked at the number in such a manner that no one else in the room could see it. Again I held it up, facing the wall and asked Eric to tap out the number. In one series of 7 tests, all calls were correct. However, in another set of 10 tests, two numbers were incorrect, each being one short of the required value. As a final test, I repeated the procedure but in this case I didn't look at the cards before holding them up to the wall. In this set, again two numbers were wrong out of a total of ten. By this time I had become certain that, whatever was responsible for the rapping sounds, they were not coming from someone on the other side of the wall. The fact that these number tests had been largely successful, and that the raps had been produced during previous visits on surfaces other than the wall, led to the conclusion that the occupants of the house next door were not responsible.

I think this falsifies the neighbour hypothesis. Indeed assuming Dr Colvin’s honesty, it makes the whole thing a bizarre episode and extremely suggestive of a discarnate intelligence, or living agent using some kind of ESP; even Dr Colvin can not see the card in the final test.

Colvin, JSPR, vol. 72; No.890; January 2008 wrote:
It is clear that in the Andover case we are not discussing the sudden unexpected movement of objects within the house; we are discussing a lengthy series of knocking sounds, capable of providing intelligent messages using a code. The idea that non-personal mechanical forces could explain these responsive phenomena cannot be seriously considered. The only normal explanation that was ever justified for further consideration was that of deception and only then during the initial phases of the investigation. During the period when rapping sounds were heard from the partitioning wall only, the deception theory was regarded as a possible explanation of events — not because there was any evidence positively to suggest this, but simply important and ultimately significant that the rapping sounds be made to emanate from objects other than the partitioning wall between the Andrews' home and that of their neighbours.

And so the case must rely on Dr Colvin’s observations. That this experience has led him to analyse the raps and resulted in the paper we have critiqued is obvious – but clearly even after 36 years Dr Colvin continues to seek answers to his experiences in Andover. Why have I placed so much emphasis on this? We must not throw out the baby with the bath water – a promising line of enquiry has closed it seems, unless future analysis shows some fault in our conclusions, but we are no nearer to understanding the original poltergeist case that led us here.

We are running out of debate: after this interesting excursion I will return to my main argument next post, and look at some findings of Dr Richard Wiseman. For those who want to do some homework, look at the papers linked here

J x